DSXchange

We recently moved all of our batch processing from SUSE Linux 11 w/ IIS v11.5 to SUSE Linux 12 w/ IIS v11.7
We've started getting these job log warnings on various different Server jobs that we never encountered before running on v.11.5
I'm working with IBM Support, but has anyone here personally dealt with successfully suppressing these warning messages?

DataStage Job 5078 Phantom 1758
uvconfig: No such file or directory
PRINTER memory segment removed
DataStage Phantom Finished

I searched forums and found really, really old responses to this issue. I understand these are purely informational messages, however we have job controls that will "Abort" if even warnings are logged. So it's causing us some headache and need to figure out how to stop these from happening. Like I mentioned before, we never saw this on v11.5 server so I'd like to know what changed or maybe what's not configured properly. Appreciate any tips, TY!

*These log messages correspond with the job log warning error message above*
/opt/IBM/InformationServer/Server/DSEngine/dsdlockd.log
Tue Aug 3 01:07:43 2021 - DeadLock Daemon started in Dead Process Cleanup Mode by pid (13175)
Tue Aug 3 01:07:43 2021 - ** Cleanup being performed for UserSig(-13174)
Tue Aug 3 01:07:43 2021 - ** Removed entry from signature table
Tue Aug 3 01:07:43 2021 - ** Removed shared memory segment for User (13174), SegNo (adebcc8a)
Tue Aug 3 03:39:42 2021 - DeadLock Daemon started in Dead Process Cleanup Mode by pid (46919)
Tue Aug 3 03:39:42 2021 - ** Cleanup being performed for UserSig(-14152)
Tue Aug 3 03:39:42 2021 - ** Removed entry from signature table
Tue Aug 3 03:39:42 2021 - ** Removed shared memory segment for User (14152), SegNo (adebc8b8)
Tue Aug 3 03:53:27 2021 - DeadLock Daemon started in Dead Process Cleanup Mode by pid (24762)
Tue Aug 3 03:53:27 2021 - ** Cleanup being performed for UserSig(-24761)
Tue Aug 3 03:53:27 2021 - ** Removed entry from signature table
Tue Aug 3 03:53:27 2021 - ** Removed shared memory segment for User (24761), SegNo (adeb9f47)
Tue Aug 3 04:16:12 2021 - DeadLock Daemon started in Dead Process Cleanup Mode by pid (11081)
Tue Aug 3 04:16:12 2021 - ** Cleanup being performed for UserSig(-11074)
Tue Aug 3 04:16:12 2021 - ** Removed entry from signature table
Tue Aug 3 04:16:12 2021 - ** Removed shared memory segment for User (11074), SegNo (adebd4be)
Tue Aug 3 05:50:49 2021 - DeadLock Daemon started in Dead Process Cleanup Mode by pid (10217)
Tue Aug 3 05:50:49 2021 - ** Cleanup being performed for UserSig(-10211)
Tue Aug 3 05:50:49 2021 - ** Removed entry from signature table
Tue Aug 3 05:50:49 2021 - ** Removed shared memory segment for User (10211), SegNo (adebd81d)
Tue Aug 3 06:22:33 2021 - DeadLock Daemon started in Dead Process Cleanup Mode by pid (10207)
Tue Aug 3 06:22:33 2021 - ** Cleanup being performed for UserSig(-10206)
Tue Aug 3 06:22:33 2021 - ** Removed entry from signature table
Tue Aug 3 06:22:33 2021 - ** Removed shared memory segment for User (10206), SegNo (adebd822)

What has your research into your missing uvconfig path? Did you validate everything in there?

I'm not sure why that's part of the warning message. I'd assume we'd run into bigger issues if that file was missing, but it's there and owned by 'dsadm' with full read/write access ==> /opt/IBM/InformationServer/Server/DSEngine/uvconfig

The logged warning message seems to be tied to this dsdlockd process running.
These are the out-of-the-box settings for the configuration we have on both the v11.5 and v11.7 server.
Any recommended changes here? Wondering if we should increase the timer and still get the logged job warning, but not as frequent?

"/opt/IBM/InformationServer/Server/DSEngine/dsdlockd.config"
start=0
timer=900
res=0
log=
recordlocks=0

look inside uvconfig. Look at the paths defined in there, perhaps one is missing...

The only directories defined in unconfig are these and they all exist
UVSPOOL /tmp
UVTEMP /tmp
CSHDISPATCH /bin/csh
SHDISPATCH /bin/sh

BUT ... this whole time I thought that error meant uvconfig could not be found, but you're right it's actually uvconfig reporting that some other file or directory could not be found. Assuming it's referring to the file getting deleted by the automated deadlock cleanup process. We could disable this and I think the errors would stop occuring, but I don't know the long-term effects of disabling this. Bigger question, why would the deadlock process remove shared memory segments in use? I also don't understand how this is running if we have "start=0"

"/opt/IBM/InformationServer/Server/DSEngine/dsdlockd.config"
start=0
timer=900
res=0
log=
recordlocks=0

Every connected DataStage process has a number of memory segments allocated, up to four. One of these, of which every process has its own, is called the PRINTER memory segment, an historical hangover from UniVerse when the database product managed its own spooling. Several administrative processes, including dsdlockd, will remove the PRINTER memory segment for defunct processes. These messages can safely be ignored, but I guess you should purge the dsdlockd log from time to time.